-/* $Id: rpnscan.c,v 1.3 2006-09-21 20:22:34 adam Exp $
+/* $Id: rpnscan.c,v 1.4 2006-10-29 17:18:05 adam Exp $
Copyright (C) 1995-2006
Index Data ApS
char prefix[20];
};
-static int scan_handle1(char *name, const char *info, int pos, void *client)
-{
- int len_prefix, idx;
- struct scan_info *scan_info = (struct scan_info *) client;
-
- len_prefix = strlen(scan_info->prefix);
- if (memcmp (name, scan_info->prefix, len_prefix))
- return 1;
- if (pos > 0)
- idx = scan_info->after - pos + scan_info->before;
- else
- idx = - pos - 1;
-
- /* skip special terms such as first-in-field specials */
- if (name[len_prefix] < CHR_BASE_CHAR)
- return 1;
-
- if (idx < 0)
- return 0;
- scan_info->list[idx].term = (char *)
- odr_malloc(scan_info->odr, strlen(name + len_prefix)+1);
- strcpy(scan_info->list[idx].term, name + len_prefix);
- assert (*info == sizeof(ISAM_P));
- memcpy (&scan_info->list[idx].isam_p, info+1, sizeof(ISAM_P));
- return 0;
-}
-
-static ZEBRA_RES rpn_scan_ver1(ZebraHandle zh, ODR stream, NMEM rset_nmem,
- struct rset_key_control *kc,
- Z_AttributesPlusTerm *zapt,
- int *position, int *num_entries,
- ZebraScanEntry **list,
- int *is_partial, RSET limit_set,
- int index_type, int ord_no, int *ords)
-{
- int pos = *position;
- int num = *num_entries;
- int before;
- int after;
- int i;
- struct scan_info *scan_info_array;
- char termz[IT_MAX_WORD+20];
- ZebraScanEntry *glist;
- int ptr[RPN_MAX_ORDS];
-
- before = pos-1;
- if (before < 0)
- before = 0;
- after = 1+num-pos;
- if (after < 0)
- after = 0;
- yaz_log(YLOG_DEBUG, "rpn_scan pos=%d num=%d before=%d "
- "after=%d before+after=%d",
- pos, num, before, after, before+after);
- scan_info_array = (struct scan_info *)
- odr_malloc(stream, ord_no * sizeof(*scan_info_array));
- for (i = 0; i < ord_no; i++)
- {
- int j, prefix_len = 0;
- int before_tmp = before, after_tmp = after;
- struct scan_info *scan_info = scan_info_array + i;
- struct rpn_char_map_info rcmi;
-
- rpn_char_map_prepare (zh->reg, index_type, &rcmi);
-
- scan_info->before = before;
- scan_info->after = after;
- scan_info->odr = stream;
-
- scan_info->list = (struct scan1_info_entry *)
- odr_malloc(stream, (before+after) * sizeof(*scan_info->list));
- for (j = 0; j<before+after; j++)
- scan_info->list[j].term = NULL;
-
- prefix_len += key_SU_encode (ords[i], termz + prefix_len);
- termz[prefix_len] = 0;
- strcpy(scan_info->prefix, termz);
-
- if (trans_scan_term(zh, zapt, termz+prefix_len, index_type) ==
- ZEBRA_FAIL)
- return ZEBRA_FAIL;
-
- dict_scan(zh->reg->dict, termz, &before_tmp, &after_tmp,
- scan_info, scan_handle1);
- }
- glist = (ZebraScanEntry *)
- odr_malloc(stream, (before+after)*sizeof(*glist));
-
- /* consider terms after main term */
- for (i = 0; i < ord_no; i++)
- ptr[i] = before;
-
- *is_partial = 0;
- for (i = 0; i<after; i++)
- {
- int j, j0 = -1;
- const char *mterm = NULL;
- const char *tst;
- RSET rset = 0;
- int lo = i + pos-1; /* offset in result list */
-
- /* find: j0 is the first of the minimal values */
- for (j = 0; j < ord_no; j++)
- {
- if (ptr[j] < before+after && ptr[j] >= 0 &&
- (tst = scan_info_array[j].list[ptr[j]].term) &&
- (!mterm || strcmp (tst, mterm) < 0))
- {
- j0 = j;
- mterm = tst;
- }
- }
- if (j0 == -1)
- break; /* no value found, stop */
-
- /* get result set for first one , but only if it's within bounds */
- if (lo >= 0)
- {
- /* get result set for first term */
- zebra_term_untrans_iconv(zh, stream->mem, index_type,
- &glist[lo].term, mterm);
- rset = rset_trunc(zh, &scan_info_array[j0].list[ptr[j0]].isam_p, 1,
- glist[lo].term, strlen(glist[lo].term),
- NULL, 0, zapt->term->which, rset_nmem,
- kc, kc->scope, 0, index_type, 0 /* hits_limit */,
- 0 /* term_ref_id_str */);
- }
- ptr[j0]++; /* move index for this set .. */
- /* get result set for remaining scan terms */
- for (j = j0+1; j<ord_no; j++)
- {
- if (ptr[j] < before+after && ptr[j] >= 0 &&
- (tst = scan_info_array[j].list[ptr[j]].term) &&
- !strcmp (tst, mterm))
- {
- if (lo >= 0)
- {
- RSET rsets[2];
-
- rsets[0] = rset;
- rsets[1] =
- rset_trunc(
- zh, &scan_info_array[j].list[ptr[j]].isam_p, 1,
- glist[lo].term,
- strlen(glist[lo].term), NULL, 0,
- zapt->term->which,rset_nmem,
- kc, kc->scope, 0, index_type, 0 /* hits_limit */,
- 0 /* term_ref_id_str */ );
- rset = rset_create_or(rset_nmem, kc,
- kc->scope, 0 /* termid */,
- 2, rsets);
- }
- ptr[j]++;
- }
- }
- if (lo >= 0)
- {
- zint count;
- /* merge with limit_set if given */
- if (limit_set)
- {
- RSET rsets[2];
- rsets[0] = rset;
- rsets[1] = rset_dup(limit_set);
-
- rset = rset_create_and(rset_nmem, kc, kc->scope, 2, rsets);
- }
- /* count it */
- count_set(zh, rset, &count);
- glist[lo].occurrences = count;
- rset_delete(rset);
- }
- }
- if (i < after)
- {
- *num_entries -= (after-i);
- *is_partial = 1;
- if (*num_entries < 0)
- {
- *num_entries = 0;
- return ZEBRA_OK;
- }
- }
- /* consider terms before main term */
- for (i = 0; i<ord_no; i++)
- ptr[i] = 0;
-
- for (i = 0; i<before; i++)
- {
- int j, j0 = -1;
- const char *mterm = NULL;
- const char *tst;
- RSET rset;
- int lo = before-1-i; /* offset in result list */
- zint count;
-
- for (j = 0; j <ord_no; j++)
- {
- if (ptr[j] < before && ptr[j] >= 0 &&
- (tst = scan_info_array[j].list[before-1-ptr[j]].term) &&
- (!mterm || strcmp (tst, mterm) > 0))
- {
- j0 = j;
- mterm = tst;
- }
- }
- if (j0 == -1)
- break;
-
- zebra_term_untrans_iconv(zh, stream->mem, index_type,
- &glist[lo].term, mterm);
-
- rset = rset_trunc
- (zh, &scan_info_array[j0].list[before-1-ptr[j0]].isam_p, 1,
- glist[lo].term, strlen(glist[lo].term),
- NULL, 0, zapt->term->which, rset_nmem,
- kc, kc->scope, 0, index_type, 0 /* hits_limit */,
- 0 /* term_ref_id_str */);
-
- ptr[j0]++;
-
- for (j = j0+1; j<ord_no; j++)
- {
- if (ptr[j] < before && ptr[j] >= 0 &&
- (tst = scan_info_array[j].list[before-1-ptr[j]].term) &&
- !strcmp (tst, mterm))
- {
- RSET rsets[2];
-
- rsets[0] = rset;
- rsets[1] = rset_trunc(
- zh,
- &scan_info_array[j].list[before-1-ptr[j]].isam_p, 1,
- glist[lo].term,
- strlen(glist[lo].term), NULL, 0,
- zapt->term->which, rset_nmem,
- kc, kc->scope, 0, index_type, 0 /* hits_limit */,
- 0 /* term_ref_id_str */);
- rset = rset_create_or(rset_nmem, kc,
- kc->scope, 0 /* termid */, 2, rsets);
-
- ptr[j]++;
- }
- }
- if (limit_set)
- {
- RSET rsets[2];
- rsets[0] = rset;
- rsets[1] = rset_dup(limit_set);
-
- rset = rset_create_and(rset_nmem, kc, kc->scope, 2, rsets);
- }
- count_set(zh, rset, &count);
- glist[lo].occurrences = count;
- rset_delete (rset);
- }
- i = before-i;
- if (i)
- {
- *is_partial = 1;
- *position -= i;
- *num_entries -= i;
- if (*num_entries <= 0)
- {
- *num_entries = 0;
- return ZEBRA_OK;
- }
- }
-
- *list = glist + i; /* list is set to first 'real' entry */
-
- yaz_log(YLOG_DEBUG, "position = %d, num_entries = %d",
- *position, *num_entries);
- return ZEBRA_OK;
-}
-
-
ZEBRA_RES rpn_scan(ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt,
oid_value attributeset,
int num_bases, char **basenames,
projects / idzebra-moved-to-github.git / commitdiff